This proposal is for a 16-week, placebo-controlled trial of naltrexone (50 mg orally) in 200 early problem drinkers (i.e., hazardous or harmful drinkers without significant alcohol dependence severity). The study will employ a factorial design in which the effects of schedule of medication administration (i.e., daily vs. targeted), as well as the interaction of medication and schedule of administration will be examined. Targeted administration refers to the use of medication to cope with anticipated high-risk drinking situations. The primary outcome measures will be drinking days and heavy drinking days. Secondary outcomes will include alcohol-related problems and biological measures of alcohol consumption (i.e., GGTP and CDT). The study will provide an opportunity to replicate and extend the results of a recently completed 8-week trial of targeted naltrexone in early problem drinkers. That study showed a significant advantage of naltrexone over placebo on heavy drinking days and a trend for an effect of targeted administration on daily drinking. In the proposed study, procedures will be refined, the sample size will be increased and the duration of treatment will be double that of the recently completed study. In addition, the daily monitoring of mood, desire to drink, perceived self-efficacy, and drinking behavior will make it possible to examine in depth the mechanisms by which the study variables exert their effects. Follow-up evaluations conducted at 3 and 6 months post-treatment will provide a measure of the durability of treatment effects. Careful evaluation of the study hypotheses will provide important information on the efficacy and mechanism of targeted naltrexone for early problem drinkers, who comprise a substantial proportion of the general population.